Nvidia reported fiscal second-quarter revenue of $96.2 billion for the period ended July 26, 2026, an increase of 106% from a year earlier and 18% sequentially. Data-centre revenue reached a record $89 billion, rising 117% year on year and accounting for most of the company’s growth as hyperscalers, AI laboratories, cloud providers and enterprises continued to expand accelerated-computing infrastructure.
GAAP net income increased to $59.69 billion from $26.42 billion a year earlier. Diluted GAAP earnings were $2.46 a share, while non-GAAP earnings were $2.22. Both GAAP and adjusted gross margins stood at 75%. Nvidia returned approximately $26 billion to shareholders through repurchases and cash dividends during the quarter.
For the current quarter, Nvidia expects revenue of approximately $108 billion, subject to a 2% variation. Its guidance assumes no data-centre compute revenue from China, isolating the forecast from a market affected by export controls and shifting licensing conditions. The company expects GAAP and non-GAAP gross margins of about 74%, plus or minus 50 basis points, with respective operating expenses of approximately $9.2 billion and $9 billion.
Chief executive Jensen Huang said demand was broadening beyond a small number of frontier laboratories, with new AI labs, startups, open-model developers and physical-AI systems all contributing to infrastructure requirements. Nvidia attributed the data-centre increase partly to the ramp-up of its Blackwell Ultra platform. The company is also moving into the Vera Rubin generation, which combines new GPU, CPU, networking and rack-scale components for training and inference.
The results provide a current measure of the capital flowing into the compute layer used by Indian AI developers, cloud operators and planned AI factories. Indian infrastructure projects have increasingly specified Nvidia systems for domestic capacity. AM Intelligence, for example, has placed an order covering 9,000 Vera Rubin NVL72 rack-scale systems for the first 30-megawatt phase of an AI facility under development in Hyderabad, with delivery scheduled for the first quarter of 2027.
Nvidia’s scale also affects the economics of Indian cloud and application companies that consume GPU capacity indirectly through global platforms. Continued growth in hyperscale procurement determines the availability, generation mix and pricing of compute offered through those providers, while the expansion of Nvidia’s enterprise and sovereign-AI business creates a parallel channel for locally deployed systems.
Management nevertheless identified supply availability as a constraint. Chief financial officer Colette Kress said customer forecasts indicated still higher potential demand, while Huang described pressure across the supply chain. Nvidia’s outlook therefore combines accelerating end-market requirements with continuing dependence on advanced manufacturing, memory, networking, power systems and data-centre construction capacity.
